On Wed, 29 Jan 2003 13:38:46 +0000, Daniel Lorch <[EMAIL PROTECTED]> said: > Not afraid of DNS-poisoning? If any of the above DNS would be 0wned > and would send falsified data for swinog.ch, the page could be > hijacked to point somewhere else. More servers = more risk.
Well, yes, but wouldn't one eventually notice when the different NSes give different results? (Unless they are all hacked; and that would require more effort when there are more servers - assuming they don't all run the same software). > RFC2182 [1], Section 5, is very clear about this: > More servers also increase the likelihood that one server will > be misconfigured, or malfunction, without being detected. That's true. But the difficulty of detecting a misconfiguring/ malfunctioning slave start as soon as there is more than one (or zero) of them. It gets significantly more when you start having slaves that are outside your administrative domain. But once you've gone down that path, one could argue that if you can detect misconfigurations and malfunctions in *one* externally operated slave, then you can also detect them in several; as long as you know who they are. Getting problems fixed once they have been detected is another matter, especially when you don't know the other operators well. Fortunately in this case you can at least be sure that they all read the SwiNOG mailing list! > Oh and even more :) > On the other hand, having large numbers of servers adds little > benefit, while adding costs. At the simplest, more servers cause > packets to be larger, so requiring more bandwidth. This may seem, > and realistically is, trivial. However there is a limit to the size > of a DNS packet, and causing that limit to be reached has more > serious performance implications. We're aware of that, and we're still safely below 400 bytes. -- Simon. ---------------------------------------------- [EMAIL PROTECTED] Maillist-Archive: http://www.mail-archive.com/swinog%40swinog.ch/
